Differences between revisions 7 and 8
Revision 7 as of 2011-09-20 02:45:50
Size: 1461
Editor: aly.deines
Comment:
Revision 8 as of 2011-09-20 05:23:05
Size: 2770
Editor: aly.deines
Comment:
Deletions are marked like this. Additions are marked like this.
Line 24: Line 24:

* From William: For L-series lovers: Getting the doctest coverage to 100% on this
might be a good project:

  http://code.google.com/p/purplesage/source/browse/psage/lseries/eulerprod.py

That may discover "issues" (bugs), which I would likely have to fix,
but would also be fun because one gets to come up with lots of
creative examples of L-series all over the place. Also, the top of
that file has a todo list for new features to implement -- most would
be bad projects, but one which would be good would be to make it so
the Lseries object can use Lcalc (Rubinstein's program) to compute
L-series instead of Dokchitser. This would be a good project, because
it would mainly involve thinking about the annoying mathematics
involved in going between normalizing L-series with the center of the
critical strip at 1/2 versus not doing that. Also, it is all pure
Python, so easier to get going.

Anyway, I'd say 1 could be a good project for people who know the
basics of L-series, but want to get a much more concrete feel for
them. In fact, instead of just trying to get coverage to 100%,
writing a *tutorial* for computing with L-series using that package
would be really nice. E.g., one could walk through how to find
missing information, create new L-series classes, etc.

To do list

* p-adics

* computing with chi_18

* wrapping of gauss composition (in pari: QuadClassUnit)

* KateWishList.sws

* #11697: Global minimal models over number fields with class number >= 1

  • this is in Connell and probably wouldn't take to long to get at least a python toy version
  • Sage already has this for class number 1 fields

* Open beginner tickets

* Reviewing number theory and elliptic curve tickets

* From William: For L-series lovers: Getting the doctest coverage to 100% on this might be a good project:

That may discover "issues" (bugs), which I would likely have to fix, but would also be fun because one gets to come up with lots of creative examples of L-series all over the place. Also, the top of that file has a todo list for new features to implement -- most would be bad projects, but one which would be good would be to make it so the Lseries object can use Lcalc (Rubinstein's program) to compute L-series instead of Dokchitser. This would be a good project, because it would mainly involve thinking about the annoying mathematics involved in going between normalizing L-series with the center of the critical strip at 1/2 versus not doing that. Also, it is all pure Python, so easier to get going.

Anyway, I'd say 1 could be a good project for people who know the basics of L-series, but want to get a much more concrete feel for them. In fact, instead of just trying to get coverage to 100%, writing a *tutorial* for computing with L-series using that package would be really nice. E.g., one could walk through how to find missing information, create new L-series classes, etc.

days33/todo (last edited 2011-09-20 21:34:07 by lola)